“Unfortunately, the types of people for whom such fellowships might represent the greatest departure from their everyday experience—and whose career trajectories might be most dramatically shifted given freedom from their usual constraints—are infrequently their beneficiaries.”
In an article at Public Books, Dominique J. Baker (University of Delaware) and Christopher T. Bennett (RTI International) share their research on nearly 30,000 fellowships awarded by six fellowship programs to which any individual in the arts, humanities, or social sciences can apply, including Guggenheim Fellowships, National Humanities Center Fellowships, and others.
These fellowships generally provide time and funding for work as well as “extensive opportunities for networking and collaboration.”
The authors write:
Sociological research suggests that it is the people who already have considerable money, status, or both that are more likely to win grants, fellowships, and awards, especially more prestigious honors that enhance the cultural cachet of recipients. This tendency exists due to the ways that social networks, cumulative advantages, and access to greater resources play a role in the preparation of the application, recommender solicitation, and ultimate selection process.
Among their findings:
- Winners of one of these fellowships are more likely to be winners of one of the others.
- The vast majority of these fellowship recipients have been based at a US college or university.
- The five most common university affiliations for grant recipients over the past 100 years have been Harvard, UC Berkeley, Columbia, Stanford, and Yale.
They write:
The fact that these institutional affiliations remain so dominant suggests that a series of decisions, preferences, and resource constraints tend to favor applicants with higher-status affiliations. One contributing factor, of course, is the likely prospect that the people who apply to these fellowships disproportionately work at high-status, well-resourced universities.
People who work at better-resourced universities may also very well have different track records precisely because they have more support routinely available to them (some get a portion of an assistant’s time, others get funds for students to help conduct research, etc.). For instance, universities with a substantial number of winners often have structures for sharing winning materials and established processes for reviewing each applicant’s draft application materials, streamlining the pathway to a fellowship for those at well-connected universities. Further, several institutions—such as Columbia, Notre Dame, the University of California–Los Angeles, the University of Texas at Austin, and the University of Wisconsin–Madison—even have the financial resources to match funding from certain high-priority fellowships, resulting in a world where people nominally winning the same size award can actually receive different funding amounts.
The strong relationship between the status of an applicant’s institutional affiliation and their chances of winning one of these prestigious fellowships is clear. Such outcomes can result in self-reinforcing cycles, with prospective applicants with less prestigious affiliations deterred from submitting applications in the first place or, as shown in prior research, electing to not apply again after receiving an initial rejection. These dynamics create the risk of missing pivotal opportunities to support the very types of groundbreaking new work that the fellowships were designed to help flourish.
The full article is here.
These articles are a tad bit reptitive! “Prestige and power and ability to pursue things like Philosophy is usually congregated in an elite class of sorts with the culture, status, and wealth to do these things.” I guess on the backend meritocracy really is a lie, which may be surprising to academics–but I do find it like, well, I suppose someone is pointing out the problem. I wish resources and the political climes were pointed at a solution, as controversial as those things might be.
Guess who’s not going to NIU and instead is trying to take classes in state, which. already stretches his budget?
It’s good to have evidence showing there really is a problem, though. Because otherwise, the skepticism is pretty overwhelming.
One difficulty with all of this though is that it struggles to distinguish causation from correlation. If you look at books that have been top-10 bestsellers at bookstores in Singapore, and at books that have been translated into French, you’ll find a strong correlation. That’s not because being a bestseller in Singapore is playing any direct causal role in the decision of a publisher to commission a French translation (except maybe in a few cases involving breakout hits by Singaporean authors) or because being translated into French makes a book sell well in Singapore.
In each particular case, there is some explanation for why both of these things happened, and it’s often a common cause. Sometimes it’s because the original publisher of the book is well-funded and advertises heavily around the world, causing interest both in Singaporean bookstores and in translators to French. But in other cases, it’s because some intrinsic features of the book make it appealing both to people in Singapore and to people in France.
Merely observing the correlation doesn’t tell us much at all about which of these pathways (or various other pathways) explains the correlation.
Similarly, it’s hard to tell if people from the same five universities keep winning major fellowships because the fellowship committees specifically care about the institution the person is at, or if it’s because the kind of work that the fellowship committees like is the same kind of work that hiring committees at these universities like, or something else. The former would be fairly pernicious, but for the latter, we’d want to investigate further to understand whether the features of work that appeal to both of these groups are the actual features that we think make work good, or are merely good advertising.
It’s hard to get that detail from this level of statistical observation without lots of detailed controls.
Which is why it’s good that people keep looking at the issue and gathering more data, and generating more analyses.
Well, a lack of controls, as it were, puts more skeptical reasoning into the debate as to “is meritocracy real” as opposed to stating anything about the actual thing itself.
I think my initial comment was somewhat ill-formed and thus open to all sorts of interpretation. It is further complicated by the fact that merit has different definitions (it suffers from the same problems, vis a vis egalitariamism). I think what meant to say, was–well, the kind of *naive* meritocracy where one puts a linear relationship on effort and say, talent, and then some ill defined notion of societal success–it is false. That being said, I do think soft meritocracy may hold to some degree–but there is a bevy of evidence to argue against the hard and fast, deterministic relationship case.
That is to say, institutional privelege and “talent” are not mutually exclusive.
The idea that it’s plausible that people “from the same five universities” win the major fellowships isn’t supported by intro to philosophy of science arguments, no matter how long-winded.
Kenny’s point does not rely on its being plausible that people from the same five universities win the major fellowships, and doesn’t rely on that plausibility being supported by introductory arguments in the philosophy of science.
I think this response is asking the article to do something different from what sociological analysis typically does.The point is not to isolate one specific causal pathway. The concentration of fellowship winners in a small set of elite institutions is itself evidence of a structured social pattern produced through multiple overlapping processes rather than a single cause. In that context, saying “it could be resources, prestige, hiring patterns, application support, committee preferences, or the kinds of work those institutions cultivate” is not a rebuttal. That is exactly what cumulative advantage looks like. That is also what makes this a sociological fact: a durable pattern of inequality produced through multiple interacting processes rather than one isolated mechanism.
So no, the correlation does not by itself prove that committees directly reward institutional names. But it does show that fellowship success is strongly patterned by institutional status, and that is already sociologically important. The existence of multiple plausible pathways does not empty the pattern of meaning. It is what makes the pattern a structural one.
The contents of your post are fair, and I think are sound if the post you’re responding to had been a parent post (i.e., a post responding to the article). But in fact, you are responding to a comment on a comment. I think that makes a difference:
The contents of the comment you are responding to was itself in response to the claim “It’s good to have evidence showing there really is a problem […] otherwise, the skepticism is pretty overwhelming.” So, I gather that the dialectic presupposes what you reject. I don’t think it is the case that Kenny is asking the article to do anything different.
If anything, I think your point is a complement to the parent post, which (charitably) sees cumulative advantage in this territory as already thoroughly established, such that the article is no real news/just more of the same, and so not of intrinsic interest to a wide audience. On that read of the parent post, I gather your comment agrees with the substantive point, and neither agrees nor disagrees with the marketing point.
Perhaps these funding organizations could consider lotteries: you need to meet a certain threshold, but once you do, the winners are determined by lottery. This has been tried on a small scale for various science grants. (I don’t think the funders of e.g., Guggenheim fellowships would ever go for this, but maybe a separate set of awards and grants determined by lottery…)
I mean, it depends on the specific criterion–you’re essentially kicking the problem of “inegalitarianism in funding” to “the criterion where chance plays a role”. It’s not as if the top philosophy students aren’t going to be clustered in top departments–you’re just giving a chance to whomever manages to get past the filter, which are necessarily going to include places of institutional prestige. So instead of a literally 0 chance, one can go to a not top-tier institutuion and get a 1/100th chance at receiving a fellowship. Which is a relative improvement, but ultimately, who’s to say that the proportion of students who make it pass the cutoff won’t disproportionately come from top departments?
I think, unfortunately, any strategy which leaves “meritocratic” selection intact will invariably run into the same issues. This is a feature of specialization of labor, not a bug.
There’s a big problem there though. This would be fine for those who want grants purely because they buy relief from teaching. But grants play other important motivational roles in the wider professional ecosystem that lotterisation would undermine entirely. Here are two examples, both turning on the merit-signalling function of a grant. First, I might want to apply for promotion from associate to full professor. Landing a large grant can be crucial here, but it wouldn’t be if grants were lotterised. The same holds for upwards trajectory in one’s international career: no one cares about lotterised results when screening for merit (imagine: “I managed to get a paper in Phil Review …. it was a lottery!”). And the ‘past a certain threshold, it’s a lottery’ move doesn’t rescue this, since in any given case it remains that the exclusive awarding of funding to this person rather than that one wasn’t a matter of merit.
But these things are lotteries anyway. Most papers in Phil Review would have been rejected if they’d got different referees; many other papers would have been accepted with somewhat better referee luck.
(My mental model here is that for an arbitrary paper, there are 20 appropriate referees, and the quality of the paper determines how many of those 20 would say to accept it. For some papers the number that would accept is 0, for vanishingly few it’s 20, but for many others it’s somewhere in between. It’s basically random which of those 20 actually referee the paper. If a paper would only be accepted by 2 out of 20, then it’s extremely lucky to get accepted; if it would be accepted by 18 out of 20, it’s extremely unlucky to be rejected. But both those things are possible, and for papers that would be accepted by 10-15 referees, it’s really a coin flip what happens.)
Making the randomness more explicit might lead people to treat things like grant awards and journal publications more appropriately.
Right, but the whole point of grants is to allow people to pursue important research. They might have taken on some kind of merit signalling function – but that isn’t what they are fundamentally for, and it probably shouldn’t be what they are fundamentally for.
“These dynamics create the risk of missing pivotal opportunities to support the very types of groundbreaking new work that the fellowships were designed to help flourish.”
Is that what they were designed for? It seems clear that’s not the case. I once went to a ridiculous–and overfunded–Templeton conference and there was a Templeton rep in attendance, so I asked him why they seem to always fund people who don’t need the career help instead of people who do. The rep told me it’s because the best connected people are going to be able to spread their message the loudest.
I can’t speak for the other fellowships, but what Templeton wants is not to help new groundbreaking work flourish; they want a specific image of philosophy elevated.